Metrics, Benchmark and Data Asset processing with Pipelines
New Functionalities
- Models benchmarking
- Metrics execution
- Dataset onboarding from S3 links
- Dataset processing with pipelines
Minor Improvements
- Deprecated old methods (e.g.
create_asset_from_id
andget_assets_from_page
) - Text Normalization added on benchmark
- Refactored SDK to have uniformity across all modules in I/O types and function names